E.B.R,[_ LIBRARY,/ _ T-72 Table 2 Joint Trusteeship of Pension Plans and its Impact on Market EBRI multiemployer Using QPIR plans data, is Davis's actually testimony attributable shows to legitimate that single-employer differences defined in investment benefit Joint Trusteeship of Pension Plans and its Impact on Market reported on the 1985 Form 5500 Sample Tape were classified into one of the following: participant Chart cohorts 3 shows are that generally single-employer more mature, defined as measured benefit plan by the sponsors proportion alsoof follow retirees the The second measurement proceeds essentially along the same lines as the first; Rates of Return, Ending September 30, 1989 Performance: A Further Examination Based on EBRI and Form 5500 Data L Performance: A Further Examination Based on EBRI and Form 5500 Data strategies would weaken the argument put forth that a joint trusteeship scenario would stocks, pension bonds, plans have or "other." a significantly Stocks higher include percentage all corporate of their stocks, portfolio both invested preferred in equity and in the participant population. After correcting for these demographic differences, the expected however the tendency differences of increasing in aggregate theirrates investment of return in are safer comput asseted classes within (i.e., eachbonds) of the ten as by Jack L. VanDerhei, Ph.D. A Period be unnecessarily expensive for single-employer plan sponsors. than do their multiemployer counterparts. Moreover, single-employer defined benefit plan common. Bonds include state and municipal securities as well as long-term U.S. Quarter 1 Year 3 Year a 5 Year a estimated the participant one year population loss formatures. single-employer The multiemployer defined benefit plans pension again show plans no in 1985 systematic from plan participant maturity classifications first. At that point a weighted average, based on Summary Plan Type and Indicies 89Q3 88Q4-89Q3 86Q4-89Q3 84Q4-89Q3 All Assets sponsors had a significantly lower percentage of their portfolio invested in bonds. Given government securities and long-term corporate debt instruments. the relationship joint trusteeship between mandate percentage actually of retirees increased andfrom asset $6.25 allocation billion decisions. to $6.46 billion. Thus percentage of total assets (chart 1), is computed for the ten differences. This "corrected" Introduction Single Employer DB Plans 7.4% 22.9% 16.5% 18.1% Statement on Data • SingleSingle-employer Employer DC Plans defined9.7 benefit % plans in 24.6% vest a larger percentage 15.8% of their 1portfolio 7.3% that the rate of return realized on equities over the sample period was greater than that the actual cost of imposing this structure on single-employer pension plans would appear differential aggregate rate of return of 119 basis points is actually larger than the All Multiemployer Plans 5.3% 18.9% 12.4% 15.5% in equities than multiemployer plans Joint Trusteeship of Pension Plans and its Impact on Market •Consumer Price Index 0.7% 4.3 % 4.3 % 3.4 % Results rea]ized on bonds (a result consistent with long term historical results), 2 her analysis Performance: A Further Examination Based on EBRI and Form 5500 Data to be larger than those previously reported. Differential Investment Results -- One method of measuring the impact of the uncorrected rate reported above resulting in a one year loss (based on 1985 asset levels) • Multiemployer plans have a higher percentage of retirees compared to their total In earlier testimony submitted by EBRI on the joint trusteeship of pension plans, Equity participant population than do single-employer defined benefit pension plans Investment Retu Hearings rns -- on TheJoint investment Trusteeship returns used of Pension in this analysis Plans are based on suggests that an additional $87 billion would need to be contributed to make up for the Single Employer DB Plans 11.0% 33.6% 20.9% 22.5% of $6.46 billion. tendency to invest more cautiously as the participant population matures is to compute Single Employer DC Plans 12.3% 33.9% 20.9% 23.3% author Jennifer Davis' provided estimates on the additional amount of net contributions Before the U.S. House Education and Labor Subcommittee on shortfall in investment income. • All M The ultiemploye purpose r Pla of nsthis research 11.6% is to investigate 34.1% whether 21.3% the more mature 22.5% cohort the five-year results reported in Table 2 of Davis (attached). This time horizon is the Maturity of Participant Populations -- Chart 1 suggests that the argument that the difference between two types of investment results for single-employer defined benefit Labor Management Relations S&P 500 10.7% 32.9% 18.6% 20.3% affects the investment allocation of multiemployer plans by analyzing the asset that would have been required by single-employer plan sponsors if nonjointly trusteed longestallocation Davis reported (pp. of 6-8) inmultiemployer QPIR discusses and wasseveral plans choseand nreasons to single-employer smooth for out thistemporary observed definedmarket difference benefit fluctuations plans in asset with . Conclusion Bonds multiemployer plans have, on average, older employees is indeed true. After categorizing 28 February 1990 plans assuming a joint trusteeship scenario was imposed. The implicit assumption is that similar demographic characteristics defined benefit plans had achieved the lower rate of return that was achieved by jointly Single Employer DB Plans 0.8% 11.8% 8.7% 12.6% by allocations The annualized between returns single-employer for single-employer and multiemployer defined benefit plans.plans Included over in this thisperiod discussion were Single Employer DC Plans 1.8% 17.5% 10.5% 13.7% each plan into one of ten classifications based on the ratio of retirees to total participants, the investment managers hired by the single-employer defined benefit sponsors will • In general, single-employer defined benefit plan sponsors follow the expected All Multiemployer Plans 0.9% 12.2% 8.8% 14.0% trusteed defined benefit plans over the previous seven years. Her testimony points out Jack VanDerhei was Shea the rson argument / Lehman b that: 0.9 % 11.3 % 8.1% 13.2 22.5% tendency in equitiesofand investing 12.6% in less bonds. of their Bas portfolio ed on 19853 in stocks year-end as the asset plan allocations participant for the total general investments (including cash) were aggregated and the distribution of continue to have the same performance results within their asset allocation; however the Senior Research Associate Although expected decreases in investment income from mandating a joint Source: population EBRI Quarterlymatures; Pension Inve nostsuch ment trend Report, is thievident rd quarterfor1989, the rmultiemployer evised. plans that one of the primary reasons that additional single-employer contributions would be Employee Benefit Research Institute aThree- and five-year returns are expressed as annualized rates. these plans, a rate of return for all other assets of 16.9% was imputed. total assets was computed. Chart 1 shows that, on a dollar-weighted basis, trusteeship overall assetscenario allocation for of single-employer the sponsor will pension be modified plans is toonly those one historically consideration chosen in the by bShearson Lehman Brothers Kuhn Loeb Government/Corporate Bond Index. • Single-employer defined benefit plan sponsors also follow the expected tendency required multiemployer under a jointplans trusteeship have, on scenario average, results olderfrom employees the difference with longer in asset tenure allocation of increasing their investment in safer asset classes (bonds) as the participant which prompts the board of trustees to invest in bonds to preserve multiemployer plans have much higher percentages of retirees with a median value in the multiemployer plan sponsors. current debate, some observers have discounted the estimated losses due to perceived population matures; multiemployer plans again show no systematic relationship of investment performance of the past, preserving the monies for the benefits between the two plan entities. Davis's analysis was based on EBRI's Quarterly Pension this type Demographic Information -- Since the maturity of the participant population is that will need to be paid in the near future. Single-employer plans with an 25 to 30 percent classification, as opposed to a median value in the 15 to 20 percent The first measurement uses the five-year rates of return for the three asset classes differences in the manner in which these assets are invested. The purpose of this overall younger workforce could invest more in equities which over many Investment Report (QPIR), which allows detailed breakdowns of investment results by years would even out fluctuations in the market. •centralTherefore, to the argument the more conservative explained earlier, asset allocation a proxy for decisions classifying made plans by multiemployer into various classification for single-employer plans. reported above and computes an overall rate of return for single-employer and testimony is to assist decision makers in their attempt to gauge the true financial impact plan sponsors does not appear to be due to the more mature participant plan entity (e.g., single-employer versus multiemployer plans), but provides no population categories of participant population maturity was needed. Using information on the 1985 multiemployer of such a change defined by controlling benefit plan for the asset most allocations likely source using of the differences 1985 year-end in investment asset The purpose of this testimony is to provide an empirical examination of this demographic information. This additional testimony will further examine the implications •Form When 5500 Sample the assetTap allocation e, ratiosofof multiemployer retirees to total plans participants is overlaid on were thecomput single-employer ed for all income. Asset Allocation Results as a Function of Plan Maturity -- Chart 2 allocations computed from the Form 5500 tape. This results in an "uncorrected" defined benefit plans and investment results are simulated based on historical of joint trusteeship through an analysis of Form 5500 data intended to examine the hypothesis. The results are important to the current debate because any documentation returns, these latter plans experience a loss of total assets due to the lower single-employer and multiemployer defined benefit pension plans. demonstrates that, in general, single-employer defined benefit plan sponsors follow the aggregate Similar rate toof the return testimony of 17.98% previously and submitted 16.83% forby single-employer Davis, this research and multi estimates employer the investment return motivations in asset allocation of jointly trusteed plans. that the lower rate of return experienced due to the asset allocation typical of The views expressed in this statement are solely those of the author and expected tendency of investing less of their portfolio in stocks as the plan participant asset allocations, respectively. When the difference of 117 basis points is multiplied by potential reduction in investment income for single-employer pension p!ans assuming that should not be attributed to the Employee Benefit Research Institute, its _4 _q officers, trustees, sponsors, or other staff. The Employee Benefit Research Asset Allocation -- Given the need to limit the analysis to the asset categories population matures (i.e., a higher ratio of retirees). However, no such trend is evident for the a joint QPIR trusteeship estimate of mandate 1985 year-end would result assets in a for modification single-employer of asset defined allocation benefit decisions plans, .a Institute is a non-profit, non-partisan public policy research organization. 2Based on historical results from 1926 to 1986, the average one-year rates of return on the major asset classes were: common stocks, 12.1%; small stocks, 18.2%; long-term reported in testimony submitted by Davis, all general investments (including cash) the one multiemployer year loss of $6.25 plans, billion with th iseobtained exception . of those plans with more than 40 percent of Unlike her testimony however, this methodology controls for the possibility that corporate 1jennifer stocks, Davis, 5.3%; "Joint longTrusteeship -term government of Pension bonds Plans , 4.7%; and and its U.S. Impact Treasury on Market bills, 3.5%. See Roger Ibbotson and Rex A. Sinquefield, Stocks, Bonds, Bills and Inflation: The Performance: An Analysis of EBRI Data," Statement for the Hearings on Joint Trusteeship their participants already retired. differences in asset allocations are a natural result of the fact that multiemployer plan of Pension Plans Before the U.S. House Education and Labor Subcommittee on Labor Past and the Future, Charlottesville, Virginia: Financial Analysts Research Foundation, 1987. _The reason for selecting this year is explained below. Management Relations, 21 February 1990. 3 1 4 7 6 5 EMPLOYEE BENEFIT RESEARCH INSTITUTE 2121 K Street, N%V/ Suite 600 / Washington, DC 20037-2121 Telephone 202-659-0670 FAX 202-775-6312